Susanne K. Langer, "'On a Fallacy in 'Scientific Fatalism'"

From International Journal of Ethics, Vol. 46, No. 4, July, 1936, 473-83. “Our very apprehension of the imminence of a future act changes the environment which makes the act imminent.”

June 31, 2009

On a Fallacy in “Scientific Fatalism”

Like a long-forgotten photograph by chance re-covered—some close friend of far-off days—Professor Perry’s Thought and Character of William James suddenly reminds us of that great and genial American whose vivid nature dominated an aca-demic era, who always seemed to stand in the very center of vital discussions and to see the immediate issues that were shaping themselves out of the new ideas and experiments of his time. Most of these issues, if they have not been settled, at least have paled a little, displaced by more crying problems, or perhaps obscured by some shift of emphasis, some new slant on old ideas; Monism and Pluralism no longer hold the philosophical stage, Functional Psychology has become too respectable to cause any commotion, even Optimism and Pessimism now-adays do not force our choice between them. Revealed Religion and Evolution have fought out their enmity, and the former has retreated to its last stronghold in the hills of a southerly state. Like a faded photograph, James’s impassioned letters now show the contours of those philosophical chimaeras, as they looked in their young and untried strength.

But one of the questions that seriously agitated James, though little discussed by men of letters at present, is still alive and vital to many thoughtful minds: that is the question of determinism versus true freedom of action. It is a perennial problem; for it is not only a theoretical affair, but a moral issue; once the philosophers have raised it, the laity cannot forget it. It was raised long before there were professional thinkers. The “insolubilium” it presented to James’s generation was merely a new version of a Sphynxian riddle. But every option presents itself in its peculiar style until it is either resolved or circumvented, and this one has met with neither fate as yet. It is to us, as it was to our fathers and teachers, a living problem—the Dilemma of Determin-ism. So, despite its age, I may be pardoned for rousing this sleeping dog once more, for the purpose of finally dispatching it.

The doctrine of determinism, in its philosophic form, is a modern version of belief in Fate. On the basis of this identification a veritable war has been waged between those who, in the interests of science, welcomed the idea of a closed causal system, and those who, from moral considerations, found it abhorrent. As usual, assent and denial were given before anyone made a logical analysis of the concepts involved, or traced the actual implications of the dogma that is supposed (both by its supporters and its haters) to re-establish Fatalism in the modern world. Yet these concepts, besides being exciting, are also very interesting from a logical standpoint. The connection between causal determinism and fatalism is not simple—in fact, they have at one time stood in contradiction to each other—and their identification rests on a genuine, howbeit somewhat “technical,” mistake. A demon-stration of this momentous misconception dispels the phantom Fate, and shows us once more—if we care to draw the moral—how heedlessly the “will to believe” outruns logical inference, and jumps at the most vigorously beckoning conclusion.

It may be well, before challenging the view that determinism implies fatalism, to consider the latter in its classical form.¹ In ancient mythology, the destinies of certain men were laid down by the mystic agency of a god; and struggle as they might, these men must consummate their assigned triumphs or sufferings, though the end might be reached by unpredictable paths. In the Christian doctrine of predestination the same fatalistic element prevails: it matters not how hard the soul may struggle that is initially condemned, or how low the elect may fall through his own guilt; the conclusion of every career is written in the stars before ever the race is run.

This belief in the omnipotence of destiny has always been countered by an equally primitive philosophy of individual action: of responsibility, justice, personal initiative, in short, of prevision and purpose within a purely causal, indifferent universe. As a man sows, so shall he reap. The very struggle of a doomed hero against his fate expresses his unbelief in the absoluteness of doom. The reason why the spectator, knowing the end to be inevitable, does not regard the struggler merely as a fool, is that he sympathizes with the philosophy of effective action even while he accepts the philosophy of fate. There is a strange conflict of two doctrines that seem to be equally fundamental: the belief that man is a puppet in the hands of higher powers, and the belief that his future is “Karma,” a function of his own deeds, determined wisely or unwisely, for better or worse, by his own decisions.

Ancient mythology and Christian mysticism gra-dually yielded to the conception of a thoroughgoing causal order, wherein the power of Words, the agency of charms and curses, have no place whatever. The universe of our scientific era is a huge network of causal relations, wherein each term, i.e., each physical event, is connected directly with its next neighbors: determined by events immediately preceding, and itself the origin of the terms which immediately follow. In such a world a man’s actions of today determine his fortunes for tomorrow. By understanding the nature of causal connection, by learning the rules of the cosmic game, he can exploit those relationships, he can play his hand in that game. No evil star, no malevolent deity presides over his life; no high destiny or heavenly crown is his birthright. He must make his bed as he would lie in it. The philosophy of causation, which reasons from item to item, has defeated the mystical metaphysics of a younger age, which interpreted present and future as the fulfilment of a divine plan—a preordained drama, wherein men were merely actors whose parts were written in advance—a sham battle, like the conflict between Ahura Mazda and Ahriman, which must turn against the latter, whatever may occur between the beginning and the end. In the cosmology of science, every occurrence matters; every event creates a new situation; and any claim to clairvoyance, beyond a natural knowledge of given causes, is humbug and deception.

A world of causal relations is necessary, indeed, for intelligent action, and consequently is a prerequi-site for the existence of any aim, intent, or responsi-bility. But there is a joker in this deck, nonetheless: the very thoroughness of our belief in causation is the joker. For, by this doctrine, every act we perform has not only an effect, but a cause as well; it is itself the effect of a cause, which in turn is a consequence of an earlier cause, and so on ad infinitum. What we do today determines what shall happen tomorrow, but since causation is regressive, our deed is by no means the ultimate cause; all the causes of our act are causes of its consequences. There is an endless chain of causes, wherein each link connects any cause in its past with any consequent in its future, and from given causes there can be only one set of effects. If all causes up to a given cross-section of the world’s events were known, all consequences would be unequivocally determined. So we do, indeed, suffer the consequences of our own acts, but we have acted only as the events of our past have predestined us to act. The upshot of scientific philosophy, then, is that responsibility becomes just as meaningless here as in a doom-driven tragedy; our struggle against heredity is as vain as the hero’s fight against fate; we are once more the puppets in a show, the innocent dupes of destiny.

This is the doctrine of determinism, which is generally regarded as a modern version of fatalism. Not only for demigods and heroes, but for every one of us, the future is established from all eternity, and from it there is no escape by wit, or luck, or heavenly intervention. Only, it is not a god, a mystic Will, or a bad fairy that spells our fate; it is Nature itself that is in conflict with our ideals and ambitions and our dream of freedom.

Whatever step we take, whatever we do, say, or even think, a world-old decree of Nature stands behind the act. Each word and thought, each breath that we draw, realizes an appointed occasion. For, in the great world-order, it has causes that could have had no other effect than just this one, and anyone who had known those causes could have predicted the personal act which is their inevitable consequent.

“Anyone who had known those causes”; this clause conveys a crucial assumption in the philoso-phy which may be termed “scientific fatalism”: the assumption that there is a theoretically knowable collection of causes for any act. In the mystical order of fate and fulfilment, there is a simple correlation between one event of the past—the act of divine or demonic Will that creates the destiny by fiat—and one event of the future, the dramatic climax which completes it. Whatever may be the source of the oracle that declares such a predetermined fate, the connection between the spell which is the cause and the disaster or triumph that is its effect is a simple relation between two terms. The nature of this relation is mystical, for it does not postulate a causal chain wherein every link follows necessarily and unequivocally from a first cause. The links are vari-able; the end may come about by one means or another; if there be interference with one line of ap-proach, another will be taken. The only certain con-sequence of the first event is the last. When this relationship has been declared to us, we can grasp its logical structure: the end follows simply—though we know not why from the beginning. The hero may know his fate, or not know it, his knowledge has no bearing at all on the final act. Other people may know it, but their knowledge has no causal con-nection with his doom. At most, it may alter the way this doom is accomplished. Had Oedipus not been exposed in the forest, he would not have met his father as a stranger on the road; but he would presu-mably have slain him in civil strife, or by some gymnastic accident, or in his cups. The foreknow-ledge on the part of his elders could alter only the manner, not the nature, of his destiny. The cause of his mischance was solely a work of the Fates. If the decree of these mystic beings was known, as it was supposed to be upon the word of the oracle, then the future event which it determined could be known just as surely: for here was one cause and one effect, and nothing else was relevant to the occurrence of the latter, than the fact that this cause was given.

It is different with scientific determinations. Surely, every act has a cause; and if the total state of the universe at any time before the occurrence of an act could be known, the act could be theoretically foreseen. But if there were such knowledge on the part of any human being, this knowledge itself would constitute an item in the “total state of the universe,” and would alter the conditions of which the act in question was supposed to be a result. Let us say, then, that what is to be known is the state of the universe, including our knowledge of the future act and its relation to the given state. But this complex, if it be known, again becomes augmented by the knowledge of which it is the object. Such a “total state of the universe” is what Whitehead and Russell have called an “illegitimate totality,” a whole which cannot be theoretically constructed.² Present knowledge of the future is itself a cause of events in the future; therefore it cannot be knowledge of all the causes that operate upon the future. Quite aside from the human impossibility of knowing more than a negligible amount about the state of the universe at any time, even a hypothetical supermind could not know that total state, because such a total would have to include the knowledge itself. In short, there is no such totality.

There are certain events which can be predicted with fair accuracy, because we have learned that only a certain class of previous events is causally relevant to them, and that our thoughts and feelings, our knowledge or ignorance, are not in this class. Notably in physical science the relevant antecedents of an event may be known, and the knowledge of them add nothing to them, so the event may be predicted. Yet every scientific prediction is made with the tacit reservation: “Other things being equal.” The expression “other things” refers to the immense body of relevant facts which is steady and familiar enough to be presupposed without explicit mention. There is a constant environment wherein the causal connection takes place. This environment is itself a complex of causes for any event that occurs in it; the slightest change in it creates a new causal nexus, and stands out as a definite new agency.

Now, in the case of personal activities, although we have undoubtedly a perfectly good causal pro-gression, wherein each member is unequivocally determined by certain preceding members, it is not true that the determining complex may be known, for in the causation of personal acts this knowledge is not itself irrelevant, as it is in physics. The environ-ment of the causal process is changed by a know-ledge of causal connections; the knowledge itself destroys the original situation. Even as we think and learn about the consequences of our present activities, we are altering the sum total of those very activities. To collect the premises, the relevant causes, of our own future is like carrying water in a sieve; there is no steady environment wherein any given cause may be said with certainty to entail one definite effect, so long as that effect is in the future. The sort of prediction that rests on the understand-ing: “Other things being equal,” is unattainable in ethics and social science, because other things are never equal. Our very apprehension of the immi-nence of a future act changes the environment which makes the act imminent.

Only in so far as our knowledge is not itself a relevant factor, can future effects of present human situations be foreseen; that means that at best we can foresee developments along very general, broad lines, with plenty of leeway for “chance variations.” That is why we can predict social events only with the sort of accuracy that belongs to statistical calcula-tions, never with the precision of a laboratory experiment: “If this, then that, other things being equal.” If by the predictability of an event we mean its unfailing consequence upon known causes, then we must admit that personal acts are not only practically unpredictable, because of the immense complexity and variety of their causes, but theoretically as well, because “all the causes of an act,” before the act itself has taken place, form an “illegitimate totality.”

But this does not mean that acts are not causally produced, that they spring from chance, caprice, or nothing at all. Every act undoubtedly has a gapless family tree of responsible ancestors, and is unambig-uously determined by them. But this detailed, complete, and flawlessly rational determination is not accessible to our view except in retrospect, since any access to it in advance of its completion would destroy it.

“Determinism” is the assumption that every event has immediate causes through which it may be completely understood. This appears to be a tenable thesis and, for all scientific purposes, an indispen-sable one. But the supposed implication that, if an event is thus determined by its immediate antece-dents, it must be predictable from them, rests on a fallacy, and the fallacy in its turn rests on a hasty generalization from physical science, where predicta-bility does happen to go with determinateness. In the sphere of human activity it does not. There the future is necessarily obscure, although the past might theoretically be understood in every detail.

The sting of “Determinism” lies in the notion that the future is really predictable,³ that only our ignor-ance hides it from us, and that somewhere—in the mind of God, perhaps—it is already known. Since causality is transitive, the “ultimate cause” of any act may be traced back to the causes of its causes, etc., and we may choose at random any “totality” of facts in the remote past as the starting-point for predicting any act in the future. But in truth the “totality” of cumulative causes breaks up at exactly the point which is, for the knower, the present; for here his knowledge enters in as a fact, and makes the “totality” impossible.

It is a short step from the belief that the future is predictable by knowledge of a remote past to the belief that the future is peculiarly determined by some pre-eminent moment of the world’s past history (this being simply the “totality of facts” we happen to have chosen); and by a figure of speech, to regard any future act as “decreed” in that pre-eminent moment. This is the line of argument whereby determinism has become identified with the doctrine of scientific fatalism. But all that the old and the new concept of fatalism really have in common is the notion of “doom,” the notion that some future act, known or unknown to the agent of it, is somewhere already entertained as inevitable. The fact that, in the fatalistic drama, a man may know his fate and struggle against it, though his knowledge and his struggle are not causally relevant to it at all, whereas in a deterministic universe the knowing and the struggle are part of the immediate, relevant environment, and determine the future just as much as they, in turn, were determined by their past—this fact is overlooked in drawing the analogy. Yet this indifference of intervening events is the essence of true fatalism. Determinism merely maintains that what we will do tomorrow is just what we will do tomorrow, and nothing else, and that if we knew how we were going to do it, we would know what it was going to be. This is really not a very radical or debatable proposition. The thesis of classical fatalism, on the other hand, is that we know certain acts are going to be performed tomorrow; how they will come about is obscure and indifferent. Their causal origin is in a single past event and operates in advance of the natural order, forcing that order into compliance with the mystic connection. Truly, all that pure determinism and fatalism hold in common is the notion that past and future are causally connected, so that the future may be predicted from the past; and in a completely causal universe, the latter half of that contention breaks down for the case of human activities.

A purely retrospective determinism loses all dramatic interest. It escapes between the horns of the dilemma which William James constructed for it,⁴ and rests in the prosaic safe haven of common sense. Everything a man does could be understood if we had enough scientific insight; that is really all it claims. The supposed consequence, that it makes no difference whether we exert ourselves or not, belongs to fatalism, for in a genuine determinism every exertion has some effect, and just what this will be depends on certain attendant causes, and is not knowable. Now, if determinism does not entail predictability of the future, there is no pragmatic difference between it and its alternative, indetermin-ism⁵; since even an indeterminist would hardly be ready to ascribe a complete lack of cause or motive to acts which are accomplished, and maintain that they did not happen “somehow,” i.e., that they happened, but happened in no way at all. The choice between an indeterminate future and a determinate unpredictable future is really what James himself has called a “dead option”; his own desire “that things not yet revealed to our knowledge may really in themselves be ambiguous,”⁶ is in violation of all pragmatic principles, for “ambiguous” could have no meaning for him except in relation to knowability, and things “really in themselves” are absolutistic chimeras. What can true ambiguity of the future mean, upon the pragmatist theory of truth, but genuine radical unpredictability? In a thoroughgoing determinism, we do not even have to assume this unpredictability by any “hazard of faith”; it is demonstrably there.

The revolt against determinism is really a senti-mental revolt against scientific fatalism, with which it has become fused in the philosophical imagination. We can certainly no longer accept James’s statement that “Fortunately, no ambiguities hang about this word or about its opposite, indeterminism . . . . Their cold mathematical sound has no sentimental associ-ations that can bribe our partiality in advance.”⁷ Two pages after this praise of impartiality, he speaks of “the deterministic sentiment,” and tells us that “What makes us . . . . determinists or indeterminists, is at bottom always some sentiment like this.”

The fact is that the very essay from which I quote these remarks has been enough to endow determinism, for a generation after, with all the terrors of fatalism: predestination, the council of the gods, inescapable fortune, doom. But what is Fate without the Oracle? What is Fortune without the fortune-teller, or Doom without a decree? The predictability of the future, the notion that it is already accomplished for some mind, human or divine, that our ignorance of it is merely human limitation, makes us feel like fools of heaven, puppets in a divine comedy or tragedy. But any power of prevision is limited to a proper part of the determinate world, namely, the realm of mechanical events, facts. So, since the “total sum of causes of a future act” cannot be constructed, the melancholy determinist knows no more than his sanguine brother, the indeterminist. And if the god have a scientific secret, he need not guard it in his holy bosom lest the Sybil betray it; for it is beyond logic and language, in the limbo of the Inconceivable.

¹ The relation between these two concepts was recently discussed, and very clearly defined, by E. V. McGilvary in his article “Freedom and Necessity in Human Affairs,” which appeared in this journal for July, 1935.

² Cf. Principia mathematica, Vol. I, chap. ii, for the original statement of the fallacy of types; or, for a somewhat simplified version, R. M. Eaton’s General Logic.

³ Cf. William James, “The Dilemma of Determinism,” in The Will To Believe (New York, 1897), p. 152: “If we are determinists, we talk about the infallibility with which we can predict one another’s conduct; while if we are indeterminists, we lay great stress on the fact that it is just because we cannot foretell one another’s conduct . . . . that life is so intensely anxious and hazardous a game.”

⁵ The identity of determinateness with predictability is taken for granted by Professor McGilvary, when he says (op. cit., p. 384) “. . . the smallest events in the physical world are matters of chance in the literal sense of the word. The exact movements of an electron can no more be predicted by the physicist than the exact date of a man’s death can be foretold by an actuary.”

On a Fallacy in “Scientific Fatalism”

Langer main page